This application represents a request for support to conduct research integrating developmental theory regarding nicotine exposure and sensitivity with state-of-the-art statistical methods aimed at defining unique population subgroups at risk for nicotine dependence. To date, risk research focused on the role of depression and alcohol use in the emergence of smoking behavior and nicotine dependence has largely been confined to efforts intended to evaluate increased probability of smoking exposure. A complimentary or alternate explanation for the association is that depression and/or alcohol use may signal increased sensitivity to nicotine dependence at similar levels of use. Stimulated by findings obtained through current grant support, the goal for the next five years is to apply both traditional and state-of-the-art statistical methods to the analysis of longitudinal data that includes assessment of DSM-IV nicotine dependence and the widest array of measures of smoking exposure available. The following specific aims will be addressed: 1) Describe individual differences in the patterning (e.g. quantity, frequency, timing and duration) of smoking behavior associated with the emergence of nicotine dependence. 2) Evaluate the role of depression and alcohol use/disorders in promoting increased exposure to smoking and /or increased sensitivity to nicotine at similar levels of use. 3) Identify prominent and unique risk/exposure pathways that predict the emergence of nicotine dependence for population subgroups. 4) Determine the attributable risk associated with each of the prominent risk/exposure pathways in order to evaluate the potential for reduction of smoking behavior in the population. Existing data will be utilized including adolescent and young adult samples. Multiple statistical techniques for evaluating the presence of population subgroups at risk for nicotine dependence at varying levels of exposure will be used including latent class analysis (LCA) and classification and regression tree (CART) analysis. These group-based approaches will be complimented with traditional regression techniques and Receiver Operator Characteristic (ROC) analysis in order to evaluate the unique and complimentary contributions of variable-centered and group-based methods for evaluating individual variability in the etiology of nicotine dependence. We expect that the completion of the proposed studies will substantially advance our understanding of smoking etiology that will inform the targets, timing and content of intervention for those individuals most sensitive to the development of nicotine dependence. PUBLIC HEALTH RELEVANCE This research will provide a more detailed understanding of the nature and onset of nicotine dependence in adolescents and young adults. The identification of individuals with high susceptibility to nicotine dependence will inform more targeted efforts for further reducing smoking prevalence.